Related Links

• Who Values Access to College? (working paper version of "Stocks and Skill Acquisition)
• Artificial Intelligence and Technological Unemployment
• What is the Impact of a Major Unconventional Monetary Policy Intervention?

Transcript

Tim Sablik: Hi, podcast listeners. It's our 200th episode of Speaking of the Economy. Over the years, I've been joined by many of our researchers to discuss a range of economic issues that matter to our communities and the Fed.

To celebrate today's milestone, I'm stepping out of the studio and paying a visit to our economists to hear what they're working on right now and how it connects to the Richmond Fed's mission to better understand the economy. So, come on a walk with me through the Research department.

[Sound of footsteps and door opening, then chatter]

Sablik: Are you ready or do you need a chance to ...

Nicholas Trachter: I'm ready.

Sablik: So, I guess the first [question is] can you introduce yourself?

Trachter: Sure. My name is Nicholas Trachter. I'm a research advisor here at the Fed and I work mostly in firm dynamics, so understanding how firms grow.

Sablik: What's one thing that you're working on right now that you're really excited about?

Trachter: Yeah, I'm pretty excited about a new project that I'm working with some co-authors — one of them is also at the Fed, Chen Yeh — that is trying to understand how firms became so large.

Something that's well known in the literature is that the average firm is pretty large [and] getting larger. But the average plant size is pretty much constant, meaning that it seems that these firms are growing large by adding establishments. You can add establishments by either building them yourself or finding somebody else in your industry that is running good establishments, and you go and buy them.

What we see in the data is that a big chunk of how firms are so large is that these large firms are snatching the good establishments from other firms that are smaller. We find that a big chunk of growing large is the ability to find what to buy in the market and make it more successful.

Sablik: What about this research is really interesting to you, and sticking to the Richmond Fed more broadly?

Trachter: To me, I find it very interesting because I really like understanding how the economy works. I try to write models of these things and it's a difficult process. It really prompts us to think about this with theory, and that I find very interesting.

Probably an answer to why firms are so large I think is important. And, at the Fed, it's always a question whether if seeing large firms is an issue, is there's a reason for market power or is the fundamental reason why firms are large is productivity?

[Sound of footsteps]

Sablik: Alright, can you introduce yourself?

Urvi Neelakantan: I'm Urvi Neelakantan and I'm a senior policy economist at the Richmond Fed.

Sablik: Can you tell us one interesting thing that you're working on now?

Neelakantan: One of the papers I'm working on right now is called "Stocks and Skill Acquisition" and it's co-authored with Kartik Athreya, Felicia Ionescu and Ivan Vidangos, who are all Fed people.

Sablik: Kartik, formerly a Richmond Fed person, now in New York.

Neelakantan: [Laughs] That's right.

Sablik: What's interesting about that work, and what have you found out so far?

Neelakantan: I'm always interested in how different individual or household financial decisions are interconnected. In this paper, we were particularly interested in whether having access to the stock market affects the value of a college education.

The story of the paper is a familiar one. People want to smooth consumption over their life cycle. Your income is low early in life, it grows and then it goes down again as you retire — your earnings even more sharply — so you want to smooth consumption. And so, what you want to do is anticipate some of those future earnings and bring them back to the present. You also want to anticipate the loss of earnings after you retire and save for that.

This hump shape is particularly pronounced for people who successfully complete college. When they are in college, they're not really earning anything. They don't want to save because they're anticipating these high future earnings. They want to borrow. And if they're borrowing, they're not using the stock market to do that.

But once they graduate from college, their earnings have this steep profile, so they can start saving at high rates. That's when they enter the stock market. So, the stock market and a college education are compliments for those who go to college and finish.

Sablik: You could imagine that if somebody chooses not to go to college, instead of investing that money in college they could invest it in the stock market.

Neelakantan: Exactly. For that group who decides that, for whatever reason, college is not for them, in the data we see that they don't have as quite a hump shaped profile. Early in life, they actually have higher stock market participation and higher stock market investments. You don't see this consumption smoothing behavior from them because they have these pretty flat earnings. And so, for the "college is not for me" group, the stock market and college are actually not complementary, but substitutable investments.

Sablik: Why does studying the stock market and how people choose to invest or not invest relate to the Fed's job?

Neelakantan: I think the way it might connect to one of the things we do at the Fed is that we have economic education functions around the system. I just think it may be an intriguing thing for them to think about as they teach people about these big financial choices that they're making and how they might relate to one another.

[Sound of footsteps and door opening]

Russell Wong: Okay, sounds good.

Sablik: Alright. Are we sounding good, Charles? Okay.

Hello, thanks for letting us in. Can you introduce yourself?

Wong: My name is Russell Wong. I am working in the banking team. My typical research is about banking, payments, but also other things like the labor market and the housing market.

One methodology which I am using will be the search model. The search model means there is no longer a centralized market where the demand and supply meet each other. Some workers have to fight for the job. We have to hunt for our house. And then we are searching around for our mortgage. That will be a pretty good example about the search and matching friction, which I feel is super interesting.

Sablik: Does that relate to something interesting that you're working on right now?

Wong: This search and matching friction makes the market not perfect. Some are looking for a job and they cannot find it. They become unemployed. But, at the same time, some jobs become vacant. These frictions become super interesting, in my opinion, when there is some new development — for example, some new technology like AI or international developments or new policy, new regulation which people are still adapting to.

Sablik: You mentioned AI. You recently wrote a paper on that topic looking at that. Could you tell us a little bit more about that?

Wong: Yeah, it will be a nice example to highlight when we see a new technology like AI which we haven't seen before and when there is a labor market search friction, which gives us very different implications from the standard theory. You can set up an AI agent, can automate our tasks, maybe can do the podcast for us [laughs]. But essentially, [AI] makes everyone more productive, right? But it also could replace some of our jobs because [they] can be automated.

The typical theory will predict that [AI] will destroy some jobs, but it will make everyone more productive. This means it will also create more jobs. So, in the standard model, the two forces kind of cancel out each other — it destroys some jobs, creates more jobs. In the end, it's just a wash.

But that's only the case when there's no friction. When AI or technology destroys some jobs, it takes some time to find a new job, especially when they create a new kind of occupation or task. People will take time to learn about it. It takes time for industry to incorporate that into their firm.

This means the transition from job destruction to job creation won't be smooth. Especially when there are more jobs destroyed [than] jobs created, the transition will be more frictional. In this case, the two forces won't cancel out anymore. It will lead to persistent unemployment.

Sablik: What was your big finding in this paper?

Wong: Let's say that AI makes people 16 percent more productive and, in the long term, maybe generates 7 percent economic growth.

Sablik: And you're drawing those estimates from?

Wong: From other studies.

Sablik: Other micro studies, right? People that looked at a particular industry.

Wong: Right, right. But we now put them together into a macro model. At the same time, we calibrate the model to the U.S. economy and then put in AI technology and see what happens in the long run.

According to the model everyone can be, in the long run, three times more productive, so 360 something percent. But at the same time in the very, very long run, about 23 percent of the jobs will be destroyed. This means that although everyone is more productive, the job loss also can be very persistent.

[sound of footsteps, then knocking on a door and door opening, then chatter]

Sablik: Hello, can you introduce yourself, please?

Horacio Sapriza: Hi, Tim. I'm Horacio Sapriza and I'm a senior economist and policy advisor. My work focuses on the relationship between credit markets and financial markets with the real economy.

Sablik: What's one interesting thing that you're working on right now?

Sapriza: Together with colleagues at other institutions, we studied the implications of large-scale central bank and conventional monetary policy intervention on bank behavior. For this purpose, we use Mario Draghi. He was the European Central Bank president in 2012. We use his "whatever it takes" speech as an experiment to basically explore how such a major and credible announcement influenced banks' lending and risk taking.

To identify the effects of the announcement, we looked at something very specific: how euro area bank subsidiaries behaved in Mexico. By looking at these international spillovers — we're talking about links between European banks and Mexico through their participation in the Mexican market — we could isolate the effects of this policy announcement from the direct economic conditions in Europe.

We document that before the speech, struggling banks were engaged in risk shifting behavior, which means they were aggressively expanding credit to, let's say, more precarious borrowers. Then, we analyzed the implications of this announcement. What we do find is, after the speech, those euro area banks that were also operating in the Mexican market did not go on a lending spree. Instead, they actually slowed down their volume growth, raised rates, and shifted their focus to safer firms.

Sablik: How do you see this work connecting to the Fed's mission?

Sapriza: There are many points of interest here.

First, from a standpoint of policy design and credibility, the Fed frequently employs unconventional measures during stress events — for instance, forward guidance, large-scale asset purchases. This paper shows that the credibility of an announcement is just as important as the actual provision of liquidity in changing the behavior of the banks.

A second thing is that a primary concern of the Fed is to prevent contagion during periods of high stress. The study shows empirically that unconventional policies can serve as a form of implicit capitalization of the banks. This incentivized the banks to shift towards safer borrowers and reduce the systemic risk, both at home and also abroad.

The third element is related precisely to global transmission and spillovers. Many of the largest banks are headquartered in the U.S. but operate globally. What we show empirically is that Fed policies could impact the lending standards of subsidiaries of U.S. banks in foreign markets. Understanding these spillovers would be vital for the Fed's role in maintaining financial stability and coordinating with international regulators. So, even though this episode is anchored on what happened with the ECB, the lessons are general and applicable to some extent to institutions like the Fed.

The main takeaway for the Fed is if you have a sufficiently large and credible enough intervention, you can restore stability to a fragile financial system without necessarily triggering a wave of moral hazard.

Sablik: Great. Well, thanks so much. That was awesome.